Comp, limiter, EQ advice for inserts for pre-recording vocals

Jonathanthomas83

Just watched a video about recording vocal takes with compressor, limiter and EQ plugins inserted and rendered into that initial recording (sorry I don't know the technical terminology for that), which essentially mimics what a good hardware preamp and pre-processor would do. It's already started to bring my vocals to life and I'm getting happier.

Question though, how should I approach compressing and limiting vocals at this stage? Sparingly, I guess, but does anyone have any good advice on how to actually use these tools to get the best results? I mean I don't mind a lot of compression, or a little, but I assume that I don't want to be limiting much at this point? But I don't know how to judge that.

Open to advice on good plugins for this, I have the LA and 76 types, but have settled on izotope dynamics for this task as it has comp settings and limiter settings and a visual aid for comp knee allowing more control. I'll add the 76/&LA and saturation, possibly, later on, in post, for character.

Any advice appreciated, many thanks.

robinbowes

I'd suggest recording clean vocals as well, so you can always start again.

Stuckfast

This sounds a bit arse backwards to me. Record the vocals, play them back in context, and decide what processing is needed on the basis of listening to them. If it helps you to sing better to hear them with compression or EQ, just use that in the monitor path.

Jonathanthomas83

Stuckfast said:

This sounds a bit arse backwards to me. Record the vocals, play them back in context, and decide what processing is needed on the basis of listening to them. If it helps you to sing better to hear them with compression or EQ, just use that in the monitor path.

@stuckfast, thank you. I appreciate this. Then in this case, I'd like to ask how you deal with recording dynamic vocals. For example, I'm constantly having to change my input gain in order to compensate for loud parts and quiet parts...the beauty of this approach is that the compressor helps with that and allows me to have a master setting on everything and have a dynamic vocal without having to change levels etc. At least that's the theory.

Here's the video which explains why it's a good thing to do, with some actual examples of pre/post differences...

Stuckfast

I'm not disputing that your vocals might benefit from some dynamics processing -- if they are as variable in level as you describe then yes you'll definitely need to do something to even that out. Sounds like what you're doing is working for you, which is great. But you need to be careful not to fall into the trap of doing something automatically because "that's how you compress vocals". What treatment is appropriate depends on the voice, the recording, the performance, the song, the arrangement, etc. 

Incidentally a lot of engineers do "ride the gain" while recording dynamic vocals, but that's quite difficult to do when you are the singer! Another widely used approach is to set up separate mics for the loud and quiet sections.

Jonathanthomas83

Stuckfast said:

I'm not disputing that your vocals might benefit from some dynamics processing -- if they are as variable in level as you describe then yes you'll definitely need to do something to even that out. Sounds like what you're doing is working for you, which is great. But you need to be careful not to fall into the trap of doing something automatically because "that's how you compress vocals". What treatment is appropriate depends on the voice, the recording, the performance, the song, the arrangement, etc. 

Incidentally a lot of engineers do "ride the gain" while recording dynamic vocals, but that's quite difficult to do when you are the singer! Another widely used approach is to set up separate mics for the loud and quiet sections.

Thank you. I appreciate your help and time! A lot! I'm still learning all this. All I know is that I was hugely demoralised last year when my vocals sounded flat and with no life, even within context of the song and with post treatment. I think this is worth a try, at least, if nothing else, it'll help me learn what I should and shouldn't be doing and what works and what doesn't. :-)

Id' love it if you could have a listen to my recording, if possible, to give me a bit of a critique..if you have time, @Stuckfast ?

Cirrus

Nothing wrong with processing things on the way in if you're confident. The advantage is if the processing takes you a bit closer to a good finished mix, you have that much more context to base further decisions on as the production comes together. It can be more inspiring.

If I'm using my usual SM7b into a neve style preamp, I like starting with a gentle broad high lift and La2a style compression set so it's not doing anything if I'm singing quietly and hits 3 or 4 db gain reduction if I'm belting. Up to 3dB GR a good la2a style is pretty much magical, and the cool thing about it is that if you go beyond 3dB GR the release speeds up so it doesn't sound pumpy. It's pretty much idiot proof.

There's someone else I've recorded a lot who's voice is naturally brighter and seems to need very little compression, so I wouldn't bother with that processing.

It's just all down to context, experience etc etc. No definitive answer.

Danny1969

The thing to bear in mind with 24 bit recording is there's no need at all to record a high level on the way in, not like we had to in the tape days or the early 16 bit Protools convertors. Warm up, record a section of the vocal where your voice will be at it's loudest ... make sure that's at least 3dB under clipping then record the song, EQ, compress and most importantly ride the vocal fader on the mix using automation to record those movements. 

Winny_Pooh

Personally if you are going this route at all then strap in a hardware pre-amp and hardware compressor. That way the compressor can mildly fix the peaks before it clips the converter.

No need to render at an early stage, but if you have low latency feel free to track w eq plugins etc but if eq is an issue change the mic or the mic position first.

Jonathanthomas83

Cirrus said:

Nothing wrong with processing things on the way in if you're confident. The advantage is if the processing takes you a bit closer to a good finished mix, you have that much more context to base further decisions on as the production comes together. It can be more inspiring.

If I'm using my usual SM7b into a neve style preamp, I like starting with a gentle broad high lift and La2a style compression set so it's not doing anything if I'm singing quietly and hits 3 or 4 db gain reduction if I'm belting. Up to 3dB GR a good la2a style is pretty much magical, and the cool thing about it is that if you go beyond 3dB GR the release speeds up so it doesn't sound pumpy. It's pretty much idiot proof.

There's someone else I've recorded a lot who's voice is naturally brighter and seems to need very little compression, so I wouldn't bother with that processing.

It's just all down to context, experience etc etc. No definitive answer.

This is heartening to read. Not because it aligns with my current thinking, but because it shows that it's a thing that others have done, and that it's not completely wrong. Also, thank you for the practical advice regarding ideal db etc. I might try the LA, but I guess I have a brighter voice than most, so like your friend, it might not be the ideal tool. It's nice to have examples though. What LA2A plugin do you use, out of interest?

Danny1969 said:

The thing to bear in mind with 24 bit recording is there's no need at all to record a high level on the way in, not like we had to in the tape days or the early 16 bit Protools convertors. Warm up, record a section of the vocal where your voice will be at it's loudest ... make sure that's at least 3dB under clipping then record the song, EQ, compress and most importantly ride the vocal fader on the mix using automation to record those movements. 

Do you mean recording section by section and changing the levels accordingly? That's what I'm doing now but it stifles creativity and is very time consuming, I'm not 100% confident in my abilities either. I guess the 3db under clipping guideline helps with that though. And lastly, do you mean automation for riding the fader during tracking/recording/printing or post?

PeteC

I use a valve mic into a Neve style pre and then lightly compress most of my vocals with a hardware LA2A style comp. 

A little goes a long way, but I find it really depends on the song, the mix, and what volume range there is between loud and quite passages.  Even so I like the sound of bit of comp on most things on the way in!
Probably bad practice but works for me! 
You can always try something like Waves Vocal Rider in software to literally look ahead and ride the most obvious peaks.  

Jonathanthomas83

PeteC said:

A little goes a long way

I think this is key to my thinking. I'm not talking about blowing the guts out of it, but a touch seems to be helping at the moment.

PeteC said:

You can always try something like Waves Vocal Rider in software to literally look ahead and ride the most obvious peaks.  

Pete, would you recommend using this for tracking or in post?

Musicwolf

Danny1969 said:

The thing to bear in mind with 24 bit recording is there's no need at all to record a high level on the way in, not like we had to in the tape days

Absolutely this.  The hangover of pushing levels at the recording stage not only seems to effect those of us who grew up with tape, it also seems to have somehow carried across to the younger generation who have grown up with digital.

You simply do not need to run out of headroom these days.

robinbowes

PeteC said:

I use a valve mic into a Neve style pre and then lightly compress most of my vocals with a hardware LA2A style comp. 

Out of interest, and off-topic: what pre and comp do you use?

Danny1969

Jonathanthomas83 said:

Cirrus said:

Nothing wrong with processing things on the way in if you're confident. The advantage is if the processing takes you a bit closer to a good finished mix, you have that much more context to base further decisions on as the production comes together. It can be more inspiring.

If I'm using my usual SM7b into a neve style preamp, I like starting with a gentle broad high lift and La2a style compression set so it's not doing anything if I'm singing quietly and hits 3 or 4 db gain reduction if I'm belting. Up to 3dB GR a good la2a style is pretty much magical, and the cool thing about it is that if you go beyond 3dB GR the release speeds up so it doesn't sound pumpy. It's pretty much idiot proof.

There's someone else I've recorded a lot who's voice is naturally brighter and seems to need very little compression, so I wouldn't bother with that processing.

It's just all down to context, experience etc etc. No definitive answer.

This is heartening to read. Not because it aligns with my current thinking, but because it shows that it's a thing that others have done, and that it's not completely wrong. Also, thank you for the practical advice regarding ideal db etc. I might try the LA, but I guess I have a brighter voice than most, so like your friend, it might not be the ideal tool. It's nice to have examples though. What LA2A plugin do you use, out of interest?

Danny1969 said:

The thing to bear in mind with 24 bit recording is there's no need at all to record a high level on the way in, not like we had to in the tape days or the early 16 bit Protools convertors. Warm up, record a section of the vocal where your voice will be at it's loudest ... make sure that's at least 3dB under clipping then record the song, EQ, compress and most importantly ride the vocal fader on the mix using automation to record those movements. 

Do you mean recording section by section and changing the levels accordingly? That's what I'm doing now but it stifles creativity and is very time consuming, I'm not 100% confident in my abilities either. I guess the 3db under clipping guideline helps with that though. And lastly, do you mean automation for riding the fader during tracking/recording/printing or post?

No, you set your gain on your interface by recording the loudest bit of vocal and making sure it doesn't clip. Then leave the gain there, record the whole song it doesn't matter the other bits are quieter. That's solved in the mixing stage. 

You ride the vocal fader at the mixing stage, pulling it down for the loudest parts and pushing it up for quieter bits of vocal. Although the compressor will also do this too much compression can change a vocal drastically so most engineers use a combination of compression and automation. This is why the SSL desk was so popular in the eighties and nineties ... it had fader automation. 

I always ride the aux send  of the vocal channel to the verb bus as well, pushing it up during the lines and pulling it down before the last word so it's not so obvious. Again, use automation to record these movements. 

Jonathanthomas83

Danny1969 said:

Jonathanthomas83 said:

Cirrus said:

Nothing wrong with processing things on the way in if you're confident. The advantage is if the processing takes you a bit closer to a good finished mix, you have that much more context to base further decisions on as the production comes together. It can be more inspiring.

If I'm using my usual SM7b into a neve style preamp, I like starting with a gentle broad high lift and La2a style compression set so it's not doing anything if I'm singing quietly and hits 3 or 4 db gain reduction if I'm belting. Up to 3dB GR a good la2a style is pretty much magical, and the cool thing about it is that if you go beyond 3dB GR the release speeds up so it doesn't sound pumpy. It's pretty much idiot proof.

There's someone else I've recorded a lot who's voice is naturally brighter and seems to need very little compression, so I wouldn't bother with that processing.

It's just all down to context, experience etc etc. No definitive answer.

This is heartening to read. Not because it aligns with my current thinking, but because it shows that it's a thing that others have done, and that it's not completely wrong. Also, thank you for the practical advice regarding ideal db etc. I might try the LA, but I guess I have a brighter voice than most, so like your friend, it might not be the ideal tool. It's nice to have examples though. What LA2A plugin do you use, out of interest?

Danny1969 said:

The thing to bear in mind with 24 bit recording is there's no need at all to record a high level on the way in, not like we had to in the tape days or the early 16 bit Protools convertors. Warm up, record a section of the vocal where your voice will be at it's loudest ... make sure that's at least 3dB under clipping then record the song, EQ, compress and most importantly ride the vocal fader on the mix using automation to record those movements. 

Do you mean recording section by section and changing the levels accordingly? That's what I'm doing now but it stifles creativity and is very time consuming, I'm not 100% confident in my abilities either. I guess the 3db under clipping guideline helps with that though. And lastly, do you mean automation for riding the fader during tracking/recording/printing or post?

No, you set your gain on your interface by recording the loudest bit of vocal and making sure it doesn't clip. Then leave the gain there, record the whole song it doesn't matter the other bits are quieter. That's solved in the mixing stage. 

You ride the vocal fader at the mixing stage, pulling it down for the loudest parts and pushing it up for quieter bits of vocal. Although the compressor will also do this too much compression can change a vocal drastically so most engineers use a combination of compression and automation. This is why the SSL desk was so popular in the eighties and nineties ... it had fader automation. 

I always ride the aux send  of the vocal channel to the verb bus as well, pushing it up during the lines and pulling it down before the last word so it's not so obvious. Again, use automation to record these movements. 

Brilliant, thank you for this.

This thread has been hugely helpful. Thank you all. Keep the tips and nuggets coming!

PeteC

robinbowes said:

PeteC said:

I use a valve mic into a Neve style pre and then lightly compress most of my vocals with a hardware LA2A style comp. 

Out of interest, and off-topic: what pre and comp do you use?

I often use an AMS (Neve copy) 1073 I built from a kit, and a dual LA2A, also built from a kit. Or I use a REDD47 clone I made (not from a kit!) .     I'm a bit of a kit/electronics nerd and hate spending 000's if I can build something myself.   I also use a Blackface 1176 (guess what....it was a kit).  My problem is sometime knowing how to use them to best effect, having no direct professional studio experience. But I've been using them for several years and have kind of found out by trial and error work works best for me.  

PeteC

Jonathanthomas83 said:

PeteC said:

A little goes a long way

I think this is key to my thinking. I'm not talking about blowing the guts out of it, but a touch seems to be helping at the moment.

PeteC said:

You can always try something like Waves Vocal Rider in software to literally look ahead and ride the most obvious peaks.  

Pete, would you recommend using this for tracking or in post?

Post - Waves V Rider looks at the track ahead and applies compression as needed - its also quite adjustable for a one-stop solution. 

PolarityMan

You can also use a monitoring fx chain so you get the performance benefits of a bit of sweetening without having it baked into the capture. 

Cirrus

 What LA2A plugin do you use, out of interest?

IK white 2a or Black Rooster Vla-2a. Honestly not used the Black Rooster enough to decide if it's better or not, but the IK one never annoyed me over several year's regular use.

robinbowes

PeteC said:

robinbowes said:

PeteC said:

I use a valve mic into a Neve style pre and then lightly compress most of my vocals with a hardware LA2A style comp. 

Out of interest, and off-topic: what pre and comp do you use?

I often use an AMS (Neve copy) 1073 I built from a kit, and a dual LA2A, also built from a kit. Or I use a REDD47 clone I made (not from a kit!) .     I'm a bit of a kit/electronics nerd and hate spending 000's if I can build something myself.   I also use a Blackface 1176 (guess what....it was a kit).  My problem is sometime knowing how to use them to best effect, having no direct professional studio experience. But I've been using them for several years and have kind of found out by trial and error work works best for me.  

Interesting. Do you mind my asking which kits? I'm contemplating going that route myself.

R.